The goals of this program are to understand how the quinolones act in mycobacteri a and to discover ways to protect the compounds from the development of resistance. Previous work showed that structure modifications at the C-8 position of fluoroquinolones increase antibacterial activity, particularly with fluoroquinolone-resistant mutants. This feature, plus other variations in fluoroquinolone structure, will be examined to explore the hypothesis that lethal activity arises in part from the dissociation of gyrase subunits attached to cleaved DNA. Since gyrase subunit dissociation is assayed as lethal activity in the absence of protein synthesis, this work may reveal ways to improve action against nongrowing bacteria. To define how low fluoroquinolone concentrations affect the development of resistance, nongyrase resistance mutants of Mycobacterium tuberculosis, obtained through selective growth at low drug concentration, will be examined for their ability to increase the frequency at which subsequent gyrase mutants are selectively enriched. This portion of the study is expected to influence fluoroquinolone dosing strategies. In patients, M. tuberculosis develops resistance so readily that anti-tuberculosis agents are administered as combination therapies; consequently, the lethal activity of new fluoroquinolones will be examined in combination with traditional agents to identify combinations of compounds that are unlikely to have intrinsic interfering activities. Traditional agents will also be combined with C-8-methoxy fluoroquinolones in a dynamic in vitro model to examine the effect of pharmacodynamic mismatch on the development of resistance. These two aspects of the program will help optimize the use of new fluoroquinolones. To provide a clinical context for the work, isolates from New York City will be examined for susceptibility to fluoroquinolones. Comparison of isolates obtained in the early 1990s with those obtained recently will indicate whether susceptibility is being lost. Principles emerging from these in vitro studies may be generally applicable to bacterial infections for which fluoroquinolone treatment is indicated; ideas concerning resistance may also extend to other compounds in which de novo antimicrobial resistance develops in a gradual, stepwise manner.